Hydroxyethyl starch 130/0.38-0.45 versus crystalloidor albumin in patients with sepsis: systematic reviewwith meta-analysis and trial sequential analysis

OPEN ACCESS

Nicolai Haase physician 1, Anders Perner professor 1, Louise Inkeri Hennings physician 1, MartinSiegemund professor 2, Bo Lauridsen physician 1, Mik Wetterslevmedical student 1, Jørn Wetterslevchief physician 3

1Department of Intensive Care, Copenhagen University Hospital-Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark; 2Department ofAnaesthesia, Intensive Care and Rescue Medicine, Baden State Hospital AG, Baden, Switzerland; 3Copenhagen Trial Unit, Centre for ClinicalIntervention Research, Copenhagen University Hospital-Rigshospitalet, Blegdamsvej 3, DK-2200 Copenhagen, Denmark

AbstractObjective To assess the effects of fluid therapy with hydroxyethyl starch130/0.38-0.45 versus crystalloid or albumin on mortality, kidney injury,bleeding, and serious adverse events in patients with sepsis.

Design Systematic review with meta-analyses and trial sequentialanalyses of randomised clinical trials.

Data sources Cochrane Library, Medline, Embase, Biosis Previews,Science Citation Index Expanded, CINAHL, Current Controlled Trials,Clinicaltrials.gov, and Centerwatch to September 2012; hand search ofreference lists and other systematic reviews; contact with authors andrelevant pharmaceutical companies.

Study selection Eligible trials were randomised clinical trials comparinghydroxyethyl starch 130/0.38-0.45 with either crystalloid or humanalbumin in patients with sepsis. Published and unpublished trials wereincluded irrespective of language and predefined outcomes.

Data extraction Two reviewers independently assessed studies forinclusion and extracted data on methods, interventions, outcomes, andrisk of bias. Risk ratios and mean differences with 95% confidenceintervals were estimated with fixed and random effects models.

Results Nine trials that randomised 3456 patients with sepsis wereincluded. Overall, hydroxyethyl starch 130/0.38-0.45 versus crystalloidor albumin did not affect the relative risk of death (1.04, 95% confidenceinterval 0.89 to 1.22, 3414 patients, eight trials), but in the predefinedanalysis of trials with low risk of bias the relative risk of death was 1.11(1.00 to 1.23, trial sequential analysis (TSA) adjusted 95% confidenceinterval 0.95 to 1.29, 3016 patients, four trials). In the hydroxyethyl starchgroup, renal replacement therapy was used more (1.36, 1.08 to 1.72,TSA adjusted 1.03 to 1.80, 1311 patients, five trials), and the relative

risk of acute kidney injury was 1.18 (0.99 to 1.40, TSA adjusted 0.90 to1.54, 994 patients, four trials). More patients in the hydroxyethyl starchgroup were transfused with red blood cells (1.29, 1.13 to 1.48, TSAadjusted 1.10 to 1.51, 973 patients, three trials), and more patients hadserious adverse events (1.30, 1.02 to 1.67, TSA adjusted 0.93 to 1.83,1069 patients, four trials). The transfused volume of red blood cells didnot differ between the groups (mean difference 65 mL, 95% confidenceinterval −20 to 149 mL, three trials).

Conclusion In conventional meta-analyses including recent trial data,hydroxyethyl starch 130/0.38-0.45 versus crystalloid or albumin increasedthe use of renal replacement therapy and transfusion with red bloodcells, and resulted in more serious adverse events in patients with sepsis.It seems unlikely that hydroxyethyl starch 130/0.38-0.45 provides overallclinical benefit for patients with sepsis.

IntroductionColloids are used more often for resuscitation in the intensivecare unit than crystalloids. The choice of colloid variesnoticeably between countries, but worldwide hydroxyethylstarch is most commonly used and thus more used than, forexample, human albumin and gelatin.1 The use of hydroxyethylstarch is controversial as the former higher molecular weighthydroxyethyl starch 200/0.5-0.6 caused acute kidney injury intwo randomised clinical trials of patients with sepsis.2 3 Thenewer starches with molecular weights of 130 kDa andsubstitution ratios ranging from 0.38 to 0.45 have been claimedto be safer, but the data to support this are insufficient.4 Owingto the lack of data on hydroxyethyl starch 130/0.38-0.45,previous systematic reviews have been inconclusive about the

Correspondence to: N Haase nicolai.haase@rh.regionh.dk

Extra material supplied by the author (see http://www.bmj.com/content/346/bmj.f839?tab=related#webextra)

Search strategy, details of included studies, and figuresReasons for exclusion of trials

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 1 of 12

Research

RESEARCH

benefits and harms of this colloid compared with other fluids.4-8The recent publication of three large trials comparinghydroxyethyl starch 130/0.38-0.45 with crystalloids in patientswith sepsis calls for an updated systematic review to inform onthe benefits and harms of this colloid in patients with sepsis,which is highly needed as fluid alternatives are available.9-11

We assessed the effects of hydroxyethyl starch 130/0.38-0.45versus crystalloids or human albumin on all cause mortality,kidney injury, bleeding, and serious adverse events in patientswith sepsis.

MethodsThis systematic review is based on the methodologyrecommended by the Cochrane Collaboration.12 The protocolwas published in the PROSPERO register (www.crd.york.ac.uk/PROSPERO) before the literature search.

Eligibility criteriaPotentially eligible trials had to be prospective and randomised,include patients with sepsis, have one intervention group thatreceived hydroxyethyl starch 130 with substitution ratiosbetween 0.38 and 0.45 in any concentration and in any carriersolution, and have at least one other intervention group thatreceived either crystalloid or human albumin.We included trials irrespective of language, publication status,patient’s age, indication for fluid therapy, and predefinedoutcomes. If the patients with sepsis constituted a subgroup ofthe trial population, we included the trial only if therandomisation was stratified for the presence of sepsis or if thepopulation with sepsis was larger than 500 participants. We alsoincluded quasirandomised and observational studies with morethan 500 patients receiving hydroxyethyl starch 130/0.38-0.45,but evaluated these for serious adverse events only. Exclusioncriteria were studies in animals, patients without sepsis,hydroxyethyl starch products of other molecular weights orsubstitution ratios, crossover studies, and studies comparinghydroxyethyl starch with other synthetic colloid solutions.

Search strategyWe searched the Cochrane central register of controlled trials,Medline, Embase, Biosis Previews, Science Citation IndexExpanded, and Cumulative Index to Nursing and Allied HealthLiterature. As hydroxyethyl starch 130/0.38-0.45was introducedon the market in 1999 we limited the search to references from1995 or later. We also hand searched the reference lists ofincluded trials and other systematic reviews of fluid therapy forfurther trials.Unpublished trials were sought through trial registries (www.controlled-trials.com, www.clinicaltrials.gov, and www.centerwatch.com), and we contacted relevant pharmaceuticalcompanies for unpublished data. The electronic literature searchwas last updated 10 September 2012. See the supplementaryfile for details of the search, including the search string.

Study selectionTwo authors (NH, LIH, BL, or MW) independently reviewedall titles and abstracts identified in the literature search andexcluded trials that were obviously not relevant. The remainingtrials were evaluated in full text. Disagreements were resolvedwith JW.

Data extractionTwo authors (NH, LIH) independently extracted informationfrom each included trial by using a pre-made data extractionform. The extracted information included trial characteristics(single or multicentre and country), characteristics of the trialparticipants (age, sex, and disease severity), criteria for inclusionand exclusion, type of intervention (indication, dosing, duration,and comparator fluid), and outcomes.The predefined primary outcomes of this review were overallmortality and number of patients still receiving renalreplacement therapy at the maximum length of follow-up. Thepredefined secondary outcomes were the number of patientsreceiving renal replacement therapy at any time during thefollow-up period, number of patients having acute kidney injury,number of patients receiving red blood cell transfusion, totalvolume of red blood cells transfused, number of patients havinga bleeding episode, estimated blood loss, and number of patientshaving one or more serious adverse events. We contacted thecorresponding authors for data on outcomes that were notreported in their publications.Translators extracted data from all relevant non-English articles.

Risk of bias assessmentTo determine the validity of the included trials, we assessed therisk of bias as advised by the Cochrane Collaboration,12including the domains of random sequence generation, allocationconcealment, blinding, incomplete outcome data, selectiveoutcome reporting, baseline imbalance, bias due to vestedfinancial interest, and academic bias. If one or more domainswere judged as being high or unclear, we classified the trial ashaving a high risk of bias. Since the need for fluids is difficultto assess objectively, the choice to give fluid instead ofvasopressors or inotropes may depend on the expected potencyof the fluid. Thus, unblindingmay lead to systematic differencesin interventions or cointerventions between the interventiongroups, so we classified all unblinded trials as being at high riskof bias for all outcomes including mortality unless study fluidswere given in fixed doses.

Statistical analysisReviewManager 5.1.6 was used for statistical analyses, and weused the TSA program version 0.9 beta (www.ctu.dk/tsa) fortrial sequential analyses. For each included trial we calculatedthe relative risks (95% confidence intervals) for dichotomousoutcomes and risk difference (95% confidence intervals) forcontinuous outcomes, and we pooled these measures inmeta-analyses.Heterogeneity among trials was quantified with inconsistencyfactor (I2) statistics. If the I2 statistic was 0, we reported theresults from a fixed effect model. If the I2 statistic was greaterthan 0, we reported the results from both random effects andfixed effects models.Sensitivity analyses included application of continuity correctionin trials of zero events13 and exclusion of the smallest trial, thelargest trial, and trials financed by industry.We did a predefined subgroup analysis with stratification oftrials according to risk of bias. To further explore possiblereasons for a high or moderate statistical heterogeneity we didan explorative post hoc subgroup analysis stratifying trialsaccording to length of follow-up.Some authors have suggested that conventional meta-analysisshould not be trusted without further evaluation, as cumulativemeta-analyses of trials are at risk of producing random errors

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 2 of 12

RESEARCH

because of sparse data and repetitive testing of accumulatingdata.14 15 We therefore challenged the meta-analyses with theapplication of trial sequential analysis—a sensitivity analysisthat widens the confidence intervals in case the data are toosparse to draw firm conclusions. Trial sequential analysis issimilar to interim analysis in a single trial where the monitoringboundaries are used to decide whether the P value is sufficientlysmall to show the anticipated effect and whether the trial shouldbe terminated early. In the same manner, trial sequentialmonitoring boundaries can be applied to meta-analyses.14-17

Trial sequential analysis depends on the quantification of therequired information size (the meta-analysis sample size). Wecalculated a diversity, D², adjusted required information sizesince the heterogeneity adjustment with I2 underestimate therequired information size.18 We did the trial sequential analysiswith the intention to maintain an overall 5% risk of a type Ierror and a power of 80%. For the calculation of the requiredinformation size we anticipated an intervention effect of a 20%relative risk increase. For renal replacement therapy, bleedings,and serious adverse events we used an anticipated effect of 35%,since we expected a much lower event proportion for theseoutcomes. For mortality, we observed only an 11% relative riskincrease in trials with low risk of bias and used this effect insteadin the trial sequential analysis of mortality. We provide the 95%confidence intervals adjusted for sparse data and repetitivetesting, which we describe as the trial sequential analysisadjusted 95% confidence intervals.

ResultsFigure 1⇓ summarises the results of the search. Themain reasonsfor exclusion of randomised trials were that the patients did nothave sepsis and the trials evaluated a hydroxyethyl starchsolution other than hydroxyethyl starch 130/0.38-0.45 (seesupplementary table).19-50No language restrictions were applied;one paper was in Spanish, one in Japanese, four in Russian, andfour in Chinese. Overall nine trials met the inclusioncriteria.9-11 51-56 One trial was still unpublished.51 The authors ofsix trials were successfully contacted9-11 51 53 54 and data wereobtained for eight.9-11 51 53-56A Chinese researcher extracted datafrom two trials published in Chinese.55 56 All other trials werepublished in English. No observational study was identifiedwith more than 500 patients with sepsis receiving hydroxyethylstarch 130/0.38-0.45 to evaluate for adverse events.

Characteristics of trialsThe four largest trials were blinded and had long term (>28days) follow-up.9-11 51The remaining trials were either unblinded,had unclear methodology, or had shorter follow-up times (≤28days). Table 1⇓ shows the characteristics of the included trials,and table 2⇓ the observation period for each outcome.

ParticipantsThe included trials enrolled 3456 adults with sepsis on anintensive care unit. One trial included a broad spectrum ofpatients on the intensive care unit, but in this review only thesubgroup of patients with sepsis were included.10 All but twotrials included patients with both sepsis and organ failure (severesepsis).10 51 The definitions of organ failure varied slightlybetween trials, but in most included various clinical signs ofhypoperfusion as, for example, oliguria, hypotension, andincreased lactate levels. Only one trial specifically stated thatall patients had septic shock.55

InterventionsThe type of hydroxyethyl starch studied was 6% Voluven(hydroxyethyl starch 130/0.4 (range 0.38-0.45) in saline,Fresenius Kabi, Bad Homburg, Germany) in six trials,10 11 51-54

6% Tetraspan (hydroxyethyl starch 130/0.42 (range 0.40-0.44)in Ringer’s acetate, B BraunMelsungen, Melsungen, Germany)in one trial,9 and 6% hydroxyethyl starch 130/0.4 without astatement of the brand name in two trials.55 56 Two trialscompared starch with human albumin 20%,52 53 whereas theremaining trials used crystalloid as comparator. In one studytwo groups received hydroxyethyl starch 130/0.4 in isotonicsaline or hypertonic saline.56 These groups were pooled andcompared with the third group receiving Ringer’s lactate.Trial fluid was used for resuscitation in eight trials9-11 51 52 54-56

and given as fixed doses in one trial.53 The duration of theintervention varied from 24 hours to the entire stay on theintensive care unit to a maximum of 90 days. Cumulative dosesof hydroxyethyl starch ranged from 2.1 litres to 6.4 litres withno obvious relation between duration of intervention and totaldose.

Bias risk assessmentThe risk of bias could be fully judged in six trials.9-11 51 53 54 Fourof these were judged to be of low risk of bias in alldomains,9 10 51 53 the fifth was sponsored by industry and hadpotential academic bias,11 and the sixth had a high risk of biasowing to lack of blinding.54

In the remaining three trials at least one domain was judged tobe unclear, but all of these trials were judged to be of high riskof bias in other domains (table 3⇓, also see the supplementaryfile).

Clinical outcomesAll cause mortalityMortality data were obtained from eight trials including 3414patients.9-11 51 53-56 The observation period in four of these trials(3156 patients) was longer than 28 days (table 2).9-11 51 Themeta-analysis of all eight trials showed no significant differencein mortality in patients treated with hydroxyethyl starch130/0.38-0.45 compared with crystalloid or albumin (randomeffects: relative risk 1.04, 95% confidence interval 0.89 to 1.22;P=0.64; fixed effect: 1.08, 0.98 to 1.19; P=0.13; I2=37%; fig2⇓). The trial sequential analysis adjusted 95% confidenceinterval was 0.70 to 1.54 (see supplementary file). Thepredefined analysis of trials with low risk of bias showed arelative risk of 1.11 (1.00 to 1.23; P=0.05; I2=0%), but the testfor subgroup difference between trials with low versus high riskof bias was not significant (P=0.13, fig 2). Trial sequentialanalysis of trials with low risk of bias showed that 3016 of therequired information size of 6237 patients was accrued. Thecumulative z curve touched the conventional boundary for harmbut did not cross the trial sequential monitoring boundary forharm (trial sequential analysis adjusted 95% confidence intervalof trials with low risk of bias 0.95 to 1.29) (fig 3⇓). However,the z curve will need to pass through the futility area to reachthe area of benefit, leaving little chance that hydroxyethyl starchwill turn out to reduce the relative risk of death with 11% iffurther trials are conducted in patients with sepsis.The post hoc subgroup analysis according to time of follow-upshowed a significant increase in all cause mortality in trials withfollow-up for more than 28 days (relative risk 1.11, 95%confidence interval 1.01 to 1.22; P=0.04; I2=0%) versus anon-significant decrease in all cause mortality in trials with

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 3 of 12

RESEARCH

follow-up for 28 days or less (0.63, 0.35 to 1.15; P=0.13). Thetest for subgroup differences was not significant at the 5% level(P=0.07, see supplementary file). The trial sequential analysisadjusted 95% confidence intervals of trials with follow-up formore than 28 days was 0.95 to 1.29 (see supplementary file).

Renal replacement therapy at end of follow-upFive trials had data on renal replacement therapy, withobservation periods ranging from 24 hours to one year.9 11 51 53 54

The Scandinavian Starch for Severe Sepsis/Septic Shock (6S)trial reported that two patients—one in each interventiongroup—were still being treated with renal replacement therapyat the end of follow-up.9 In Basel Starch Evaluation in Sepsis(BaSES)51 no patient required renal replacement therapy afterone year, and in the trial by Guidet et al (CRYSTMAS)11 onepatient in the hydroxyethyl starch group was treated with renalreplacement therapy for more than 28 days, but it was unclearwhether this lasted until end of follow-up. These data did notundergo meta-analysis.

Renal replacement therapy at anytime duringfollow-upThe same five trials had data on the number of patients treatedwith renal replacement therapy at anytime during follow-up.One trial had zero events in three days.53 The pooled analysisshowed that patients receiving hydroxyethyl starch130/0.38-0.45 had a significantly increased risk of receivingrenal replacement therapy (relative risk 1.36, 95% confidenceinterval 1.08 to 1.72; P=0.009; I2=0%; fig 4⇓). Application ofan empirical continuity correction of 0.01 in the no event trialdid not change the result. Trial sequential analysis showed that1311 of the required information size of 1654 patients wasaccrued, but the cumulative z curve crossed the trial sequentialmonitoring boundary for harm providing firm evidence ofincreased use of renal replacement therapy in patients treatedwith hydroxyethyl starch compared with crystalloid or albumin(trial sequential analysis adjusted 95% confidence interval 1.03to 1.80) (fig 5⇓).

Acute kidney injuryAcute kidney injury was defined as a twofold increase of serumcreatinine levels during the observation period, as this wasconsistently reported in the four trials with data on kidneyfunction.9 11 53 54 The observation periods ranged from 24 hoursto the entire stay on the intensive care unit. One trial had noevents,53 and the pooled analysis of the remaining three trialsshowed a non-significant increase in the risk of acute kidneyinjury in the hydroxyethyl starch group (relative risk 1.18, 95%confidence interval 0.99 to 1.40; P=0.07; I2=0%) (seesupplementary file). Application of an empirical continuitycorrection of 0.01 in the no event trial did not change the result.The trial sequential analysis adjusted 95% confidence intervalwas 0.90 to 1.54 (see supplementary file).

Transfusions with red blood cells, bleeding, andblood lossThree trials provided data on transfusions, with observationperiods ranging from 24 hours to the entire stay on the intensivecare unit.9 11 54 The risk of being transfused with red blood cellswas significantly higher in the hydroxyethyl starch group (1.29,95% confidence interval 1.13 to 1.48; P<0.001; I2=0%) (seesupplementary file). The trial sequential analysis adjusted 95%confidence interval was 1.10 to 1.51, providing firm evidencefor an increased risk of transfusion with red blood cells if treated

with hydroxyethyl starch 130/0.38-0.45 (see supplementaryfile).The mean volume of red blood cells did not differ between thegroups (mean difference 65 mL, 95% confidence interval −20to 149 mL; P=0.13; I2=0%) (see supplementary file).The number of patients having at least one bleeding episode(relative risk 1.34, 95% confidence interval 0.81 to 2.21; P=0.26;I2=38%) and blood loss (mean difference 26 mL, −89 to 140mL; P=0.66; I2=0%) were reported in two trials (seesupplementary file).9 11

Serious adverse eventsFour trials reported serious adverse events, two of whichregistered these during the entire stay on the intensive careunit.9 11 53 54 In the 6S trial serious adverse events were restrictedto severe bleeding and severe allergic reactions,9 whereasCRYSTMAS used broad criteria.11 The last two trials did notspecify the definition of serious adverse events, and one of themhad zero events in 24 hours follow-up.54 According to the goodclinical practice guidelines by the International Conference onHarmonisation, death should count as a serious adverse eventin the analysis,57 but we were unable to get the compositeendpoint of either death or serious adverse events from morethan one trial.9

The pooled analysis of the three trials showed a significantlyincreased risk of serious adverse events with hydroxyethyl starch130/0.38-0.45 (relative risk 1.30, 95% confidence interval 1.02to 1.67; P=0.03; I2=0%) (see supplementary file). Theapplication of a continuity correction to the zero event trialneither changed the estimate nor the confidence interval. Thetrial sequential analysis adjusted 95% confidence interval was0.93 to 1.83 (see supplementary file).

DiscussionThe main finding of this systematic review was that patientsassigned to hydroxyethyl starch 130/0.38-0.45 had inconventional meta-analysis a statistically significant increasedrisk of getting renal replacement therapy, transfusion with redblood cells, and serious adverse events. The recent large, welldesigned trials showed consistent results with no statisticalheterogeneity and the findings are likely to be confirmed whenfurther data of the patients with sepsis in the Crystalloid versusHydroxyethyl Starch (CHEST) trial10 become available, sincethe hydroxyethyl starch group in this trial had more use of renalreplacement therapy and transfusion with red blood cells andmore serious adverse events.The pooled analysis of mortality showed neither benefit norharm, but trials with a low risk of bias suggested an excessmortality of 11%. In addition, our post hoc analysis of trialswith follow-up for more than 28 days showed increasedmortality. Thus the pooled analysis of mortality may beinfluenced by trials of poor quality and too short follow-up,making interpretation difficult.The sensitivity analysis with trial sequential analysis widenedthe confidence intervals of the conventional meta-analyses whendata were too sparse to draw firm conclusions. With this strictapproach the increased risk of renal replacement therapy andtransfusion with red blood cells remained statistically significant.For mortality in trials with low risk of bias and long termfollow-up, trial sequential analysis indicated a lack of statisticalsignificance for increased mortality, but also that it is unlikelythat hydroxyethyl starch will result in a relative mortality

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 4 of 12

RESEARCH

reduction of 11% if further trials are conducted in patients withsepsis.Our results are consistent with the fact that a high fraction ofhydroxyethyl starch 130/0.38-0.45 is deposited in the tissueswhere it cannot be metabolised58 and may act as a foreign bodywith long term toxic effects, which have been described in thekidney, liver, and bone marrow.59-61 In addition, the use of renalreplacement therapy has repeatedly been associated withdeath.62 63 Our findings are in alignment with the results of twosepsis trials of hydroxyethyl starch 200/0.5-0.6 on renalimpairment and late adverse effects.2 3 Thus the adverse effectsof hydroxyethyl starch may be a class effect independent ofmolecular weight and substitution ratio.Some hypothesise that bad outcome in patients treated withhydroxyethyl starch is due to inappropriate dosing, includingthe lack of predefined triggers and goals for fluid resuscitation.No data currently support this belief, as there was no suggestionof an overall favourable outcome in any trial with adequate biascontrol and follow-up—not even in the trial designed by one ofthe manufacturers of hydroxyethyl starch.11

Strengths and limitations of the reviewThe compliance with the recommendations of the CochraneCollaboration is a major strength of our systematic review. Thisincluded a prepublished protocol, an up to date extensiveliterature search with no language restrictions, independentscreening of all references by two authors, inclusion of trialsirrespective of publication and language status and reportedoutcomes, independent data extraction by two authors, bias riskassessment, and contact with the corresponding authors of theincluded trials for additional information. In addition, wereduced the risk of random error in the meta-analyses with theapplication of trial sequential analysis using predefined variablesto increase the robustness of this analysis.We excluded trials comparing hydroxyethyl starch with othersynthetic colloids that may possess the same harmful effectsand thereby mask any adverse effects of hydroxyethyl starch.64To get a clinical applicable result, we restricted the review tohydroxyethyl starch 130/0.38-0.45 as clinicians almostexclusively use these starches. Including all types ofhydroxyethyl starch in the analysis would probably have resultedin a stronger group difference instead.The post hoc subgroup analysis of mortality in trials accordingto length of follow-up might have resulted in spurious findings.In general, however, some adverse effects undoubtedly developslowly, and if the observation period is too short, such eventsmay not be captured. In the largest trials of hydroxyethyl starchin sepsis the relative risk of death increased from day 28 to day90, indicating that the observation period for mortality shouldbe longer than 28 days, and this was the rationale for oursubgroup analysis.2 9 10

We chose to include trials with either crystalloid or albuminsolutions as comparators as no adverse effects were seen withalbumin versus saline in patients with severe sepsis in a largeintensive care unit trial.65 However, most of the included trialscompared hydroxyethyl starch with a crystalloid, and this mayprevent us from drawing firm conclusions on the effects ofalbumin. Neither can this review tell whether patients other thanthose with sepsis may experience adverse effects fromhydroxyethyl starch, but the CHEST trial found increasedserious adverse events and use of renal replacement therapywith hydroxyethyl starch in a broad population of intensive careunit patients, suggesting adverse effects beyond those seen insepsis.

Additional limitations of this review are due to bias of theincluded trials, inadequate follow-up, and trials not reportingall the outcome measures. The definitions of serious adverseevents were heterogeneous, so the group difference should beinterpreted with caution. The RIFLE (Risk of renal dysfunction,Injury to the kidney, Failure of kidney function, Loss of kidneyfunction and End-stage renal disease)66 andAKIN (Acute KidneyInjury Network)67 classifications may be better measures foracute kidney injury, but we used renal replacement therapy anddoubling of creatinine levels instead as these more simpleoutcomes were more often reported.

Relation to other reviews and implication forfuture researchSeveral well conducted systematic reviews have been publishedon hydroxyethyl starch 130/0.38-0.454 8 and on hydroxyethylstarch and fluid therapy in general.5-7Owing to the previous lackof data on hydroxyethyl starch 130/0.38-0.45, these reviewshave been inconclusive about the benefit and harm ofhydroxyethyl starch 130/0.38-0.45 compared with other fluids.In comparison, this review contains data from new large trialsand applies trial sequential analysis on the results.Hydroxyethyl starch 130/0.38-0.45 is often used in the surgicalsetting and may continue despite the raised safety issues inpatients with sepsis. If use does continue, then well poweredsurgical trials are urgently needed to ensure the safety ofpatients.

ConclusionIn conventional meta-analyses including recent trial data,hydroxyethyl starch 130/0.38-0.45 versus crystalloid or albuminin patients with sepsis was associated with an increased use ofrenal replacement therapy and transfusion with red blood cellsand more serious adverse events. The pooled analysis ofmortality showed no group difference, but this analysis may beinfluenced by trials of low quality. After trial sequential analysisadjustment for sparse data and multiple updating in cumulativemeta-analysis it seems unlikely that hydroxyethyl starchprovides overall clinical benefit for patients with sepsis.

We thank Sarah Klingenberg, search coordinator for the CochraneHepato-Biliary Group, for performing the literature search; Xia Yun forevaluating references in Chinese and extracting data from two Chinesearticles included in the review; Dimitrenka Nikolova, Copenhagen TrialUnit, for evaluating four papers in Russian; Naoya Sakamoto forevaluating one paper in Japanese; and Arnaldo Dubin, Petr Svoboda,and Martin Westphal for providing data for this systematic review. Thetrial protocol was registered through PROSPERO (registration NoCRD42011001762).Contributors: NH developed the protocol, was responsible for thesearches, selected trials, extracted data, assessed the risk of bias oftrials, did the data analysis, and developed the final review. APdeveloped the protocol, analysed data, and developed the final review.LIH developed the protocol, selected trials, extracted data, assessedthe risk of bias of trials, and developed the final review. MS extracteddata, assessed the risk of bias of trials, and developed the final review.BL and MW developed the protocol, selected trials, and developed thefinal review. JW developed the initial idea for the review, developed theprotocol, selected trials, advised on statistical methods, analysed data,and developed the final review. All authors read and approved the finalmanuscript. NH and JW are the guarantors.Competing interests: All authors have completed the ICMJE uniformdisclosure form at www.icmje.org/coi_disclosure.pdf (available onrequest from the corresponding author) and declare: AP was principal

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 5 of 12

RESEARCH

What is already known on this topic

Hydroxyethyl starches (HES) with molecular weights of 130 kDa and substitution ratios ranging from 0.38 to 0.45 are the most commonlyused colloids wordwide, but their safety and efficacy have not been established in patients with sepsisOwing to lack of data, previous systematic reviews on HES 130/0.38-0.45 and on HES in general have been inconclusive about thebenefits and harms of HES compared with other fluids

What this study adds

This systematic review includes the results of four recent randomised clinical trials of HES 130/0.38-0.45 comprising more than 3000patients with sepsisThe pooled analysis of trials showed that treatment with HES increased the risk of having renal replacement therapy, red blood celltransfusion, and severe adverse reactionsIt seems unlikely therefore that HES provides overall clinical benefit for patients with sepsis

investigator for the Scandinavian Starch for Severe Sepsis/Septic Shock(6S) trial and NH and JW were members of the steering committee.The 6S trial was funded by the Danish Research Council, theRigshospitalet Research Council, and the Scandinavian Society ofAnaesthesiology and Intensive Care Medicine (the ACTA foundation).B Braun Melsungen delivered trial fluid to all sites free of charge. Neitherthe funders nor B Braun Melsungen had an influence on the protocol,trial conduct, data analyses, or reporting of the 6S trial. AP is head ofresearch in his intensive care unit, which receives research funds fromFresenius Kabi and BioPorto. B BraunMelsungen has covered his travelexpenses for presenting 6S data at the German Anaesthetic Congress2012. MS was principal investigator for the Basel starch evaluation insepsis (BaSES) trial. The BaSES trial was funded by the Departmentof Anaesthesia and Intensive Care of the University Hospital Basel.Fresenius Kabi delivered study fluids for free and paid for the packagingand blinding process. A signed contract between Fresenius Kabi andMS stated that MS was free to publish all data without influence fromFresenius Kabi. Fresenius Kabi covered travel expenses for MS’sparticipation in meetings about fluid resuscitation.Ethical approval: Not required.Data sharing: No additional data available.

1 Finfer S, Liu B, Taylor C, Bellomo R, Billot L, Cook D, et al. Resuscitation fluid use incritically ill adults: an international cross-sectional study in 391 intensive care units. CritCare 2010;14:R185.

2 Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, et al. Intensiveinsulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med2008;358:125-39.

3 Schortgen F, Lacherade JC, Bruneel F, Cattaneo I, Hemery F, Lemaire F, et al. Effectsof hydroxyethylstarch and gelatin on renal function in severe sepsis: a multicentrerandomised study. Lancet 2001;357:911-6.

4 Hartog CS, Kohl M, Reinhart K. A systematic review of third-generation hydroxyethylstarch (HES 130/0.4) in resuscitation: safety not adequately addressed. Anesth Analg2011;112:635-45.

5 Dart AB, Mutter TC, Ruth CA, Taback SP. Hydroxyethyl starch (HES) versus other fluidtherapies: effects on kidney function. Cochrane Database Syst Rev 2010;(1):CD007594.

6 Perel P, Roberts I. Colloids versus crystalloids for fluid resuscitation in critically ill patients.Cochrane Database Syst Rev 2011;(6):CD000567.

7 Bunn F, Trivedi D, Ashraf S. Colloid solutions for fluid resuscitation. Cochrane DatabaseSyst Rev 2012;(6):CD001319.

8 Gattas DJ, Dan A, Myburgh J, Billot L, Lo S, Finfer S. Fluid resuscitation with 6%hydroxyethyl starch (130/0.4) in acutely ill patients: an updated systematic review andmeta-analysis. Anesth Analg 2012;114:159-69.

9 Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, Åneman A, et al.Hydroxyethyl starch 130/0.42 versus Ringer’s acetate in severe sepsis. N Engl J Med2012;367:124-34.

10 Myburgh JA, Finfer S, Bellomo R, Billot L, Cass A, Gattas D, et al. Hydroxyethyl starchor saline for fluid resuscitation in intensive care. N Engl J Med 2012;367:1901-11.

11 Guidet B, Martinet O, Boulain T, Philippart F, Poussel JF, Maizel J, et al. Assessment ofhemodynamic efficacy and safety of 6% hydroxyethylstarch 130/0.4 vs. 0.9% NaCl fluidreplacement in patients with severe sepsis: The CRYSTMAS study.Crit Care 2012;16:R94.

12 Higgins JPT, Green S. Cochrane handbook for systematic reviews of interventions, version5.1.0 [updated March 2011]. The Cochrane Collaboration 2011. Available from www.cochrane-handbook.org; 2012.

13 Sweeting MJ, Sutton AJ, Lambert PC. What to add to nothing? Use and avoidance ofcontinuity corrections in meta-analysis of sparse data. Stat Med 2004;23:1351-75.

14 Wetterslev J, Thorlund K, Brok J, Gluud C. Trial sequential analysis may establish whenfirm evidence is reached in cumulative meta-analysis. J Clin Epidemiol 2008;61:64-75.

15 Higgins JP, Whitehead A, Simmonds M. Sequential methods for random-effectsmeta-analysis. Stat Med 2011;30:903-21.

16 Thorlund K, Imberger G, Walsh M, Chu R, Gluud C, Wetterslev J, et al. The number ofpatients and events required to limit the risk of overestimation of intervention effects inmeta-analysis—a simulation study. PLoS One 2011;6:e25491.

17 Brok J, Thorlund K, Gluud C, Wetterslev J. Trial sequential analysis reveals insufficientinformation size and potentially false positive results in many meta-analyses. J ClinEpidemiol 2008;61:763-9.

18 Wetterslev J, Thorlund K, Brok J, Gluud C. Estimating required information size byquantifying diversity in random-effects model meta-analyses. BMC Med Res Methodol2009;9:86.

19 Ando Y, Terao Y, Fukusaki M, Yamashita K, Takada M, Tanabe T, et al. Influence oflow-molecular-weight hydroxyethyl starch on microvascular permeability in patientsundergoing abdominal surgery: comparison with crystalloid. J Anesth 2008;22:391-6.

20 Wills BA, Nguyen MD, Ha TL, Dong TH, Tran TN, Le TT, et al. Comparison of three fluidsolutions for resuscitation in dengue shock syndrome. N Engl J Med 2005;353:877-89.

21 Asfar P, Kerkeni N, Labadie F, Gouëllo JP, Brenet O, Alquier P. Assessment ofhemodynamic and gastric mucosal acidosis with modified fluid versus 6% hydroxyethylstarch: a prospective, randomized study. Intens Care Med 2000;26:1282-7.

22 Boldt J, Heesen M, Welters I, Padberg W, Martin K, Hempelmann G. Does the type ofvolume therapy influence endothelial-related coagulation in the critically ill? Brit J Anaesth1995;75:740-6.

23 Boldt J, Muller M, Heesen M, Neumann K, Hempelmann GG. Influence of different volumetherapies and pentoxifylline infusion on circulating soluble adhesion molecules in criticallyill patients. Crit Care Med 1996;24:385-91.

24 Boldt J, Mueller M, Menges T, Papsdorf M, Hempelmann G. Influence of different volumetherapy regimens on regulators of the circulation in the critically ill. Brit J Anaesth1996;77:480-7.

25 Boldt J, Müller M, Mentges D, Papsdorf M, Hempelmann G. Volume therapy in the criticallyill: is there a difference? Intens Care Med 1998;24:28-36.

26 Boldt J, Suttner S, Huttner I, Kumle B, Piper S, KrumholzW. Are cost of a crystalloid-basedvolume replacement regimen lower than of a colloid-based volume replacement strategy?Infusionsther Transfusionsmed 2001;28:144-9.

27 Boldt J, Haisch G, Suttner S, Kumle B, Schellhaass A. Effects of a newmodified, balancedhydroxyethyl starch preparation (Hextend) on measures of coagulation. Brit J Anaesth2002;89:722-8.

28 Boldt J, Ducke M, Kumle B, Papsdorf M, Zurmeyer EL. Influence of different volumereplacement strategies on inflammation and endothelial activation in the elderly undergoingmajor abdominal surgery. Intens Care Med 2004;30:416-22.

29 Boldt J, Schölhorn T, Mayer J, Piper S, Suttner S. The value of an albumin-basedintravascular volume replacement strategy in elderly patients undergoing major abdominalsurgery. Anest Analg 2006;103:191-9.

30 Bulanov Alu, Gorodetskii V, Shulutko EM, Vasil’ev SA, Orel EB, Malofeev VN, et al. Effectof different colloid volume-replacing solutions on a changed hemostasis system.Anesteziologiia i Reanimatologiia 2004;Mar-Apr:25-30.

31 Chen F, Wang C, Wang Y, Lin M, Ke J, Zhang Z. Effect of infusion with different volumeexpanders on cytokines in choleystitis patients undergoing cholecystectomy.Med JWuhanUniversity 2006;27:520-2.

32 Franz A, Braeunlich P, Gamsjager T, Felfernig M, Gustorff B, Kozek-Langenecker SA.The effects of hydroxyethyl starches of varying molecular weights on platelet function.Anesth Analg 2001;92:1402-7.

33 Gondos T, Marjanek Z, Ulakcsai Z, Szabo Z, Bogar L, Karolyi M, et al. Short-termeffectiveness of different volume replacement therapies in postoperative hypovolaemicpatients. Eur J Anaesth 2010;27:794-800.

34 Haas T, Preinreich A, Oswald E, Pajk W, Berger J, Kuehbacher G, et al. Effects of albumin5% and artificial colloids on clot formation in small infants. Anaesthesia 2007;62:1000-7.

35 Harten J, Crozier JE, McCreath B, Hay A, McMillan DC, McArdle CS, et al. Effect ofintraoperative fluid optimisation on renal function in patients undergoing emergencyabdominal surgery: a randomised controlled pilot study (ISRCTN 11799696). Int J Surg2008;6:197-204.

36 Van der Heijden M, Verheij J, van Nieuw Amerongen GP, Groeneveld AB. Crystalloid orcolloid fluid loading and pulmonary permeability, edema, and injury in septic and nonsepticcritically ill patients with hypovolemia. Crit Care Med 2009;37:1275-81.

37 Jones SB, Whitten CW, Despotis GJ, Monk TG. The influence of crystalloid and colloidreplacement solutions in acute normovolemic hemodilution: a preliminary survey ofhemostatic markers. Anesth Analg 2003;96:363-8.

38 Jover JL, García JP, Martínez C, Espí A, Gregori E, Almagro J. [Hydroxyethyl starch toprotect renal function in laparoscopic surgery].Rev Esp Anestesiol Reanim 2009;56:27-30.

39 Kumle B, Boldt J, Piper S, Schmidt C, Suttner S, Salopek S. The influence of differentintravascular volume replacement regimens on renal function in the elderly. Anesth Analg1999;89:1124-30.

40 Li F, Sun H, Han XD. The effect of different fluids on early fluid resuscitation in septicshock. Zhongguo Wei Zhong Bing Ji Jin Yi Xue 2008;20:472-5.

41 Mukhtar A, Aboulfetouh F, Obayah G, Salah M, Emam M, Khater Y, et al. The safety ofmodern hydroxyethyl starch in living donor liver transplantation: a comparison with humanalbumin. Anest Analg 2009;109:924-30.

42 Pape L, Ahlenstiel T, Ehrich J. Equivalent efficacy of HES 130/0.4 and albumin infusionsin children. Pediatr Nephrol 2008;23:1630.

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 6 of 12

RESEARCH

43 Protsenko DN, Leiderman IN, Grigor’ev EV, Kokarev EA, Levit AL, Gel’fand BR. Evaluationof the effectiveness and safety of synthetic colloid solutions in the treatment of severeabdominal sepsis: a randomized comparative study. Anesteziol Reanim 2009;9-13.

44 Schramko AA, Suojaranta-Ylinen RT, Kuitunen AH, Raivio PM, Kukkonen SI, Niemi TT.Comparison of the effect of 6% hydroxyethyl starch and gelatine on cardiac and strokevolume index: a randomized, controlled trial after cardiac surgery. Perfusion-UK2010;25:283-91.

45 Kim DH, Beez M, Kulkarni H. Comparative hemodynamic evaluation of hydroxyethylstarch (HES 130/0.4) and albumin in high-risk surgical patients. Transfus Altern TransfusMed 2005;7:79.

46 Reine PA, Kongsgaard UE, Andersen A, Thøgersen AK, Olsen H. Infusion of albuminattenuates changes in serum protein binding of drugs in surgical patients compared withvolume replacement with HAES. Acta Anaesth Scand 2008;52:406-12.

47 Trof RJ, Sukul SP, Twisk JW, Girbes AR, Groeneveld AB. Greater cardiac response ofcolloid than saline fluid loading in septic and non-septic critically ill patients with clinicalhypovolaemia. Intensive Care Med 2010;36:697-701.

48 Veneman TF, Nijhuis JO, Woittiez AJJ. Human albumin and starch administration incritically ill patients: a prospective randomized clinical trial.Wien Klin Wochenschr2004;116:305-9.

49 Verheij J, van Lingen A, Raijmakers PGHM, Rijnsburger ER, Veerman DP, Wisselink W,et al. Effect of fluid loading with saline or colloids on pulmonary permeability, oedemaand lung injury score after cardiac and major vascular surgery. Brit J Anaesth2006;96:21-30.

50 Du XJ, Hu WM, Xia Q, Huang ZW, Chen GY, Jin XD, et al. Hydroxyethyl starchresuscitation reduces the risk of intra-abdominal hypertension in severe acute pancreatitis.Pancreas 2011;40:1220-5.

51 Basel Starch Evaluation in Sepsis (BaSES). Last verified August 2011. 2012. http://clinicaltrials.gov/ct2/show/NCT00273728.

52 Palumbo D, Servillo G, D’Amato L, Volpe ML, Capogrosso G, De Robertis E, et al. Theeffects of hydroxyethyl starch solution in critically ill patients. Minerva Anestesiol2006;72:655-64.

53 Dolecek M, Svoboda P, Kantorová I, Scheer P, Sas I, Bíbrová J, et al. Therapeuticinfluence of 20% albumin versus 6% hydroxyethylstarch on extravascular lung water inseptic patients: a randomized controlled trial. Hepato-gastroenterol 2009;56:1622-8.

54 Dubin A, Pozo MO, Casabella CA, Murias G, Moseinco MC, Kanoore Edul VS, et al.Comparison of 6% hydroxyethyl starch 130/0.4 and saline solution for resuscitation of themicrocirculation during the early goal-directed therapy of septic patients. J Crit Care2010;25:659.e1-659.e8.

55 Lv J, Zhao H-Y, Liu F, An Y-Z. The influence of lactate Ringer solution versus hydroxyethylstarch on coagulation and fibrinolytic system in patients with septic shock. Zhongguo WeiZhong Bing Ji Jin Yi Xue 2012;24:38-41.

56 Zhu G-C, Quan Z-Y, Shao Y-S, Zhao J-G, Zhang Y-T. The study of hypertonic saline andhydroxyethyl starch treating severe sepsis. Zhongguo Wei Zhong Bing Ji Jin Yi Xue2011;23:150-3.

57 ICH Steering Committee. International conference on harmonisation of technicalrequirements for registration of pharmaceuticals for human use. ICH harmonised tripartiteguideline for good clinical practice, 3rd ed. Brookwood Medical Publications;1998.

58 Bellmann R, Feistritzer C, Wiedermann CJ. Effect of molecular weight and substitutionon tissue uptake of hydroxyethyl starch: a meta-analysis of clinical studies. ClinPharmacokinet 2012;51:225-36.

59 Pillebout E, Nochy D, Hill G, Conti F, Antoine C, Calmus Y, et al. Renal histopathologicallesions after orthotopic liver transplantation (OLT). Am J Transplant 2005;5:1120-9.

60 Christidis C, Mal F, Ramos J, Senejoux A, Callard P, Navarro R, et al. Worsening ofhepatic dysfunction as a consequence of repeated hydroxyethylstarch infusions. J Hepatol2001;35:726-32.

61 Schmidt-Hieber M, Loddenkemper C, Schwartz S, Arntz G, Thiel E, Notter M. Hydropslysosomalis generalisatus—an underestimated side effect of hydroxyethyl starch therapy?Eur J Haematol 2006;77:83-5.

62 Clermont G, Acker CG, Angus DC, Sirio CA, Pinsky MR, Johnson JP. Renal failure in theICU: comparison of the impact of acute renal failure and end-stage renal disease on ICUoutcomes. Kidney Int 2002;62:986-96.

63 Bagshaw SM, Uchino S, Bellomo R, Morimatsu H, Morgera S, Schetz M, et al. Septicacute kidney injury in critically ill patients: clinical characteristics and outcomes. Clin JAm Soc Nephrol 2007;2:431-9.

64 Thomas-Rueddel DO, Vlasakov V, Reinhart K, Jaeschke R, Rueddel H, Hutagalung R,et al. Safety of gelatin for volume resuscitation—a systematic review and meta-analysis.Intensive Care Med 2012;38:1134-42.

65 Finfer S, McEvoy S, Bellomo R, McArthur C, Myburgh J, Norton R. Impact of albumincompared to saline on organ function andmortality of patients with severe sepsis. IntensiveCare Med 2011;37:86-96.

66 Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, and the ADQI workgroup. Acuterenal failure—definition, outcome measures, animal models, fluid therapy and informationtechnology needs: the Second International Consensus Conference of the Acute DialysisQuality Initiative (ADQI) Group. Crit Care 2004;8:R204-12.

67 Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, et al. Acute KidneyInjury Network: report of an initiative to improve outcomes in acute kidney injury.Crit Care2007;11:R31.

Accepted: 30 January 2013

Cite this as: BMJ 2013;346:f839This is an open-access article distributed under the terms of the Creative CommonsAttribution Non-commercial License, which permits use, distribution, and reproduction inany medium, provided the original work is properly cited, the use is non commercial andis otherwise in compliance with the license. See: http://creativecommons.org/licenses/by-nc/2.0/ and http://creativecommons.org/licenses/by-nc/2.0/legalcode.

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 7 of 12

RESEARCH

Tables

Table 1| Characteristics of included studies

Contactwith

Total dose ofHES

InterventionperiodComparator

HESsolution

Indicationfor

interventionDiagnostic

group

No ofintervention

groupsBlinding

Centrestatus,setting

No ofpatientsTrial

authorssuccessful

YesMedian 3000(IQR 1507-5100)

ICU stay.Maximum 90days

Ringer’sacetate

6%Tetraspan*

ResuscitationSevere sepsis2YesMulticentre,ICU

8046S9

YesMedian 3775(IQR 2018-6347)

5 daysIsotonicsaline

6%Voluven†

ResuscitationSepsis2YesTwo ICUsin onehospital

241BaSES51

YesMean 2104 (SD850‡)

ICU stay.Maximum 90days

Isotonicsaline

6%Voluven†

ResuscitationSepsis2YesMulticentre,ICU

1937CHEST10

YesMean 2615 (SD1499)

4 daysIsotonicsaline

6%Voluven†

ResuscitationSevere sepsis2YesMulticentre,ICU

196CRYSTMAS11

Yes4×250mL/day in3 days

3 daysAlbumin 20%6%Voluven†

Fixed doseSevere sepsis2NoSingle, ICU56Dolecek200953

YesMean 2610 (SD885)

24 hoursIsotonicsaline

6%Voluven†

ResuscitationSepsis andtissuehypoperfusion

2NoMulticentre,ICU

25Dubin 201054

NoMean 2770 (SD590)

24 hoursRinger’slactate

UnclearResuscitationSeptic shock2UnclearSingle, ICU42Lv 201255

NoNo informationon doses

UnclearAlbumin 20%6%Voluven†

Maintenanceof pulmonarycapillarywedgepressure

Severe sepsis2NoSingle, ICU20Palumbo200652

NoHES+hypertonicsaline group:mean 5475 (SD209), HESgroup: mean6383 (SD 287)

24 hoursRinger’slactate

6% HES130/0.4(unclearbrand)

ResuscitationSevere sepsis3NoSingle, ICU135Zhu 201156

HES=hydroxyethyl starch; ICU=intensive care unit; IQR=interquartile range; SD=standard deviation.*6% hydroxyethyl starch 130/0.42 in Ringer’s acetate (B Braun Melsungen, Melsungen, Germany).†6% hydroxyethyl starch 130/0.4 in saline (Fresenius Kabi, Bad Homburg, Germany).‡Only reported for first four days.

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 8 of 12

RESEARCH

Table 2| Observation period for outcomes

Serious adverse eventsBleeding and blood lossRed blood celltransfusionAcute kidney injuryRenal replacement therapyMortalityTrial

ICUICUICUICU90 days90 days6S9

————1 year1 yearBaSES51

—————90 daysCHEST10

ICU4/8 daysICUICUICU90 daysCRYSTMAS11

72 hours——72 hours72 hours28 daysDolecek 200953

24 hours—24 hours24 hours24 hours28 daysDubin 201054

—————Unclear*Lv 201255

——————Palumbo 200652

—————24 hoursZhu 201156

ICU=intensive care unit.*Death in hospital or ICU, although not specifically stated.

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 9 of 12

RESEARCH

Table 3| Risk of bias

Academic biasVested financial

interestsBaselineimbalance

Selectiveoutcomereporting

Incompleteoutcome dataBlinding

Allocationconcealment

RandomsequencegenerationTrial

LowLowLowLowLowLowLowLow6S9

LowLowLowLowLowLowLowLowBaSES51

LowLowLowLowLowLowLowLowCHEST10

HighHighLowLowLowLowLowLowCRYSTMAS11

LowLowLowLowLowLowLowLowDolecek 200953

LowLowLowLowLowHighLowLowDubin 201054

UnclearUnclearLowLowUnclearUnclearHighLowLv 201255

LowUnclearLowHighLowUnclearHighUnclearPalumbo 200652

UnclearLowLowLowUnclearHighHighUnclearZhu 201156

See supplementary file to support judgment.

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 10 of 12

RESEARCH

Figures

Fig 1 Flow of papers through review. Each of the 32 excluded randomised clinical trials may have more than one reasonfor exclusion

Fig 2 Forest plot of all cause mortality in relation to risk of bias in trials. Size of squares for risk ratio reflects weight of trialin pooled analyses. Horizontal bars represent 95% confidence intervals

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 11 of 12

RESEARCH

Fig 3 Trial sequential analysis of mortality in four trials with low risk of bias. A diversity adjusted information size of 6237patients was calculated using α=0.05 (two sided), β=0.20 (power 80%), D2=0%, an anticipated relative risk increase of 11%,and an event proportion of 30% in the control arm. The blue cumulative z curve was constructed using a random effectsmodel

Fig 4 Forest plot of renal replacement therapy. Size of squares for risk ratio reflects weight of trial in pooled analyses.Horizontal bars represent 95% confidence intervals

Fig 5 Trial sequential analysis of renal replacement therapy. A diversity adjusted information size of 1654 patients wascalculated using α=0.05 (two sided), β=0.20 (power 80%), D2=0%, an anticipated relative risk increase of 35% and an eventproportion of 15% in the control arm. The blue cumulative z curve was constructed using a fixed effects model. Trials withno events were included in the analysis with an empirical continuity correction of 0.01

No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe: http://www.bmj.com/subscribe

BMJ 2013;346:f839 doi: 10.1136/bmj.f839 (Published 15 February 2013) Page 12 of 12

RESEARCH